This is an automated email from the ASF dual-hosted git repository.
github-actions[bot] pushed a commit to branch asf-staging
in repository https://gitbox.apache.org/repos/asf/datafusion-site.git
The following commit(s) were added to refs/heads/asf-staging by this push:
new 63d5e1c Commit build products
63d5e1c is described below
commit 63d5e1cd761d18cf74ef62680baa9417ceca1ca3
Author: Build Pelican (action) <[email protected]>
AuthorDate: Tue Aug 4 17:27:45 2026 +0000
Commit build products
---
blog/2026/08/07/datafusion-comet-1.0.0/index.html | 34 +++++++++++++++++++++--
blog/feeds/all-en.atom.xml | 32 +++++++++++++++++++--
blog/feeds/blog.atom.xml | 32 +++++++++++++++++++--
blog/feeds/pmc.atom.xml | 32 +++++++++++++++++++--
4 files changed, 118 insertions(+), 12 deletions(-)
diff --git a/blog/2026/08/07/datafusion-comet-1.0.0/index.html
b/blog/2026/08/07/datafusion-comet-1.0.0/index.html
index a995c1b..9a96499 100644
--- a/blog/2026/08/07/datafusion-comet-1.0.0/index.html
+++ b/blog/2026/08/07/datafusion-comet-1.0.0/index.html
@@ -54,6 +54,7 @@
<aside class="toc-container d-md-none mb-2">
<div class="toc"><span class="toctitle">Contents</span><ul>
+<li><a href="#the-road-to-10">The Road to 1.0</a></li>
<li><a href="#what-10-means">What 1.0 Means</a></li>
<li><a href="#native-expression-performance">Native Expression
Performance</a></li>
<li><a href="#correctness">Correctness</a></li>
@@ -97,10 +98,36 @@ limitations under the License.
improved performance and efficiency without requiring any code changes.</p>
<p>This release covers roughly six weeks of development since 0.17.0 and
consists of 244 commits from 23
contributors. See the <a
href="https://github.com/apache/datafusion-comet/blob/main/docs/source/changelog/1.0.0.md";>change
log</a> for the full list of changes.</p>
+<h2 id="the-road-to-10">The Road to 1.0<a class="headerlink"
href="#the-road-to-10" title="Permanent link">¶</a></h2>
+<p>Comet was <a
href="https://datafusion.apache.org/blog/2024/03/06/comet-donation/";>donated</a>
to the Apache DataFusion project in March 2024 and cut its first release,
0.1.0, five
+months later with 15 data types, 13 operators, 106 expressions, and a "modest
performance speedup." The
+nineteen releases between then and now cover a lot of ground:</p>
+<ul>
+<li><strong>Query coverage went from a handful of operators to the shape of a
real Spark query.</strong> Native
+ SortMergeJoin, HashJoin, and BroadcastHashJoin landed early; native columnar
and native shuffle,
+ broadcast nested loop joins, native window functions, native sampling, and
mixed partial/final aggregation
+ followed. Supported expressions grew from 106 at 0.1.0 to 404 in 1.0, and
the introduction of the JVM
+ codegen dispatcher in 0.17.0 gave Comet a way to keep unsupported
expressions Arrow-native by running
+ Spark's own generated code inside the pipeline rather than falling back to
row-based execution.</li>
+<li><strong>Spark support broadened.</strong> The 0.1.0 release targeted Spark
3.3, 3.4, and 3.5, with experimental 4.0. The
+ 1.0 line drops 3.3, adds 4.1, and ships an experimental 4.2 profile. ANSI
semantics — on by default in
+ Spark 4 — moved from partial to a supported default across the natively
implemented surface.</li>
+<li><strong>The ecosystem story filled in.</strong> Native Iceberg support
arrived in 0.10.0 and has been extended through
+ 1.11 and format V3, native Parquet writes and CSV reads landed as
experimental features and matured,
+ Azure joined S3 on the native cloud path, and 1.0 adds experimental
accelerated PyArrow UDFs alongside the
+ Java and Scala UDF support that shipped in 0.17.0.</li>
+<li><strong>Correctness rigor grew alongside the surface area.</strong> Early
releases relied on fuzz testing to surface
+ divergences; later releases added a full Spark SQL test-suite run against
every supported Spark version,
+ followed by AI-assisted expression audits comparing Comet's behavior against
every supported Spark version
+ edge case by edge case. 1.0 ships with every known open correctness gap
documented in the compatibility
+ guide.</li>
+<li><strong>Performance moved from "modest speedup" to workload-level
wins.</strong> TPC-DS at 1TB has become the reference
+ workload, with successive releases removing FFI round trips, caching Parquet
metadata, shrinking plan
+ serialization, and tightening the shuffle write path. 0.17.0 alone was ~9%
faster than 0.16.0 on TPC-DS
+ 1TB, and the per-expression optimization work in 1.0 continues that
trajectory.</li>
+</ul>
<h2 id="what-10-means">What 1.0 Means<a class="headerlink"
href="#what-10-means" title="Permanent link">¶</a></h2>
-<p>The 1.0.0 release is the culmination of more than two years of work since
the project was <a
href="https://datafusion.apache.org/blog/2024/03/06/comet-donation/";>donated</a>
as an
-Apache DataFusion subproject in March 2024, and is less about any single new
feature than about the
-accumulated maturity of the project:</p>
+<p>Beyond the accumulated features, 1.0 is where the project commits to being
something you can build on:</p>
<ul>
<li><strong>Broad Spark coverage.</strong> Comet supports Apache Spark 3.4.3,
3.5.9, 4.0.4, and 4.1.3 out of the same
codebase, with dedicated Maven profiles, shim sources, and CI matrices for
each, plus an experimental
@@ -328,6 +355,7 @@ to get up and running, then point Comet at your existing
Spark workloads and see
</div> </div>
<aside class="toc-container d-none d-md-block col-md-4 col-xl-3 ms-xl-2">
<div class="toc"><span class="toctitle">Contents</span><ul>
+<li><a href="#the-road-to-10">The Road to 1.0</a></li>
<li><a href="#what-10-means">What 1.0 Means</a></li>
<li><a href="#native-expression-performance">Native Expression
Performance</a></li>
<li><a href="#correctness">Correctness</a></li>
diff --git a/blog/feeds/all-en.atom.xml b/blog/feeds/all-en.atom.xml
index 6575186..76c6bd6 100644
--- a/blog/feeds/all-en.atom.xml
+++ b/blog/feeds/all-en.atom.xml
@@ -45,10 +45,36 @@ limitations under the License.
improved performance and efficiency without requiring any code
changes.</p>
<p>This release covers roughly six weeks of development since 0.17.0 and
consists of 244 commits from 23
contributors. See the <a
href="https://github.com/apache/datafusion-comet/blob/main/docs/source/changelog/1.0.0.md">change
log</a> for the full list of changes.</p>
+<h2 id="the-road-to-10">The Road to 1.0<a class="headerlink"
href="#the-road-to-10" title="Permanent link">¶</a></h2>
+<p>Comet was <a
href="https://datafusion.apache.org/blog/2024/03/06/comet-donation/">donated</a>
to the Apache DataFusion project in March 2024 and cut its first release,
0.1.0, five
+months later with 15 data types, 13 operators, 106 expressions, and a "modest
performance speedup." The
+nineteen releases between then and now cover a lot of ground:</p>
+<ul>
+<li><strong>Query coverage went from a handful of operators to the
shape of a real Spark query.</strong> Native
+ SortMergeJoin, HashJoin, and BroadcastHashJoin landed early; native columnar
and native shuffle,
+ broadcast nested loop joins, native window functions, native sampling, and
mixed partial/final aggregation
+ followed. Supported expressions grew from 106 at 0.1.0 to 404 in 1.0, and
the introduction of the JVM
+ codegen dispatcher in 0.17.0 gave Comet a way to keep unsupported
expressions Arrow-native by running
+ Spark's own generated code inside the pipeline rather than falling back to
row-based execution.</li>
+<li><strong>Spark support broadened.</strong> The 0.1.0
release targeted Spark 3.3, 3.4, and 3.5, with experimental 4.0. The
+ 1.0 line drops 3.3, adds 4.1, and ships an experimental 4.2 profile. ANSI
semantics — on by default in
+ Spark 4 — moved from partial to a supported default across the natively
implemented surface.</li>
+<li><strong>The ecosystem story filled in.</strong> Native
Iceberg support arrived in 0.10.0 and has been extended through
+ 1.11 and format V3, native Parquet writes and CSV reads landed as
experimental features and matured,
+ Azure joined S3 on the native cloud path, and 1.0 adds experimental
accelerated PyArrow UDFs alongside the
+ Java and Scala UDF support that shipped in 0.17.0.</li>
+<li><strong>Correctness rigor grew alongside the surface
area.</strong> Early releases relied on fuzz testing to surface
+ divergences; later releases added a full Spark SQL test-suite run against
every supported Spark version,
+ followed by AI-assisted expression audits comparing Comet's behavior against
every supported Spark version
+ edge case by edge case. 1.0 ships with every known open correctness gap
documented in the compatibility
+ guide.</li>
+<li><strong>Performance moved from "modest speedup" to
workload-level wins.</strong> TPC-DS at 1TB has become the reference
+ workload, with successive releases removing FFI round trips, caching Parquet
metadata, shrinking plan
+ serialization, and tightening the shuffle write path. 0.17.0 alone was ~9%
faster than 0.16.0 on TPC-DS
+ 1TB, and the per-expression optimization work in 1.0 continues that
trajectory.</li>
+</ul>
<h2 id="what-10-means">What 1.0 Means<a class="headerlink"
href="#what-10-means" title="Permanent link">¶</a></h2>
-<p>The 1.0.0 release is the culmination of more than two years of work
since the project was <a
href="https://datafusion.apache.org/blog/2024/03/06/comet-donation/">donated</a>
as an
-Apache DataFusion subproject in March 2024, and is less about any single new
feature than about the
-accumulated maturity of the project:</p>
+<p>Beyond the accumulated features, 1.0 is where the project commits to
being something you can build on:</p>
<ul>
<li><strong>Broad Spark coverage.</strong> Comet supports
Apache Spark 3.4.3, 3.5.9, 4.0.4, and 4.1.3 out of the same
codebase, with dedicated Maven profiles, shim sources, and CI matrices for
each, plus an experimental
diff --git a/blog/feeds/blog.atom.xml b/blog/feeds/blog.atom.xml
index 6981629..281afbb 100644
--- a/blog/feeds/blog.atom.xml
+++ b/blog/feeds/blog.atom.xml
@@ -45,10 +45,36 @@ limitations under the License.
improved performance and efficiency without requiring any code
changes.</p>
<p>This release covers roughly six weeks of development since 0.17.0 and
consists of 244 commits from 23
contributors. See the <a
href="https://github.com/apache/datafusion-comet/blob/main/docs/source/changelog/1.0.0.md">change
log</a> for the full list of changes.</p>
+<h2 id="the-road-to-10">The Road to 1.0<a class="headerlink"
href="#the-road-to-10" title="Permanent link">¶</a></h2>
+<p>Comet was <a
href="https://datafusion.apache.org/blog/2024/03/06/comet-donation/">donated</a>
to the Apache DataFusion project in March 2024 and cut its first release,
0.1.0, five
+months later with 15 data types, 13 operators, 106 expressions, and a "modest
performance speedup." The
+nineteen releases between then and now cover a lot of ground:</p>
+<ul>
+<li><strong>Query coverage went from a handful of operators to the
shape of a real Spark query.</strong> Native
+ SortMergeJoin, HashJoin, and BroadcastHashJoin landed early; native columnar
and native shuffle,
+ broadcast nested loop joins, native window functions, native sampling, and
mixed partial/final aggregation
+ followed. Supported expressions grew from 106 at 0.1.0 to 404 in 1.0, and
the introduction of the JVM
+ codegen dispatcher in 0.17.0 gave Comet a way to keep unsupported
expressions Arrow-native by running
+ Spark's own generated code inside the pipeline rather than falling back to
row-based execution.</li>
+<li><strong>Spark support broadened.</strong> The 0.1.0
release targeted Spark 3.3, 3.4, and 3.5, with experimental 4.0. The
+ 1.0 line drops 3.3, adds 4.1, and ships an experimental 4.2 profile. ANSI
semantics — on by default in
+ Spark 4 — moved from partial to a supported default across the natively
implemented surface.</li>
+<li><strong>The ecosystem story filled in.</strong> Native
Iceberg support arrived in 0.10.0 and has been extended through
+ 1.11 and format V3, native Parquet writes and CSV reads landed as
experimental features and matured,
+ Azure joined S3 on the native cloud path, and 1.0 adds experimental
accelerated PyArrow UDFs alongside the
+ Java and Scala UDF support that shipped in 0.17.0.</li>
+<li><strong>Correctness rigor grew alongside the surface
area.</strong> Early releases relied on fuzz testing to surface
+ divergences; later releases added a full Spark SQL test-suite run against
every supported Spark version,
+ followed by AI-assisted expression audits comparing Comet's behavior against
every supported Spark version
+ edge case by edge case. 1.0 ships with every known open correctness gap
documented in the compatibility
+ guide.</li>
+<li><strong>Performance moved from "modest speedup" to
workload-level wins.</strong> TPC-DS at 1TB has become the reference
+ workload, with successive releases removing FFI round trips, caching Parquet
metadata, shrinking plan
+ serialization, and tightening the shuffle write path. 0.17.0 alone was ~9%
faster than 0.16.0 on TPC-DS
+ 1TB, and the per-expression optimization work in 1.0 continues that
trajectory.</li>
+</ul>
<h2 id="what-10-means">What 1.0 Means<a class="headerlink"
href="#what-10-means" title="Permanent link">¶</a></h2>
-<p>The 1.0.0 release is the culmination of more than two years of work
since the project was <a
href="https://datafusion.apache.org/blog/2024/03/06/comet-donation/">donated</a>
as an
-Apache DataFusion subproject in March 2024, and is less about any single new
feature than about the
-accumulated maturity of the project:</p>
+<p>Beyond the accumulated features, 1.0 is where the project commits to
being something you can build on:</p>
<ul>
<li><strong>Broad Spark coverage.</strong> Comet supports
Apache Spark 3.4.3, 3.5.9, 4.0.4, and 4.1.3 out of the same
codebase, with dedicated Maven profiles, shim sources, and CI matrices for
each, plus an experimental
diff --git a/blog/feeds/pmc.atom.xml b/blog/feeds/pmc.atom.xml
index 2a24dc0..3bda8b4 100644
--- a/blog/feeds/pmc.atom.xml
+++ b/blog/feeds/pmc.atom.xml
@@ -45,10 +45,36 @@ limitations under the License.
improved performance and efficiency without requiring any code
changes.</p>
<p>This release covers roughly six weeks of development since 0.17.0 and
consists of 244 commits from 23
contributors. See the <a
href="https://github.com/apache/datafusion-comet/blob/main/docs/source/changelog/1.0.0.md">change
log</a> for the full list of changes.</p>
+<h2 id="the-road-to-10">The Road to 1.0<a class="headerlink"
href="#the-road-to-10" title="Permanent link">¶</a></h2>
+<p>Comet was <a
href="https://datafusion.apache.org/blog/2024/03/06/comet-donation/">donated</a>
to the Apache DataFusion project in March 2024 and cut its first release,
0.1.0, five
+months later with 15 data types, 13 operators, 106 expressions, and a "modest
performance speedup." The
+nineteen releases between then and now cover a lot of ground:</p>
+<ul>
+<li><strong>Query coverage went from a handful of operators to the
shape of a real Spark query.</strong> Native
+ SortMergeJoin, HashJoin, and BroadcastHashJoin landed early; native columnar
and native shuffle,
+ broadcast nested loop joins, native window functions, native sampling, and
mixed partial/final aggregation
+ followed. Supported expressions grew from 106 at 0.1.0 to 404 in 1.0, and
the introduction of the JVM
+ codegen dispatcher in 0.17.0 gave Comet a way to keep unsupported
expressions Arrow-native by running
+ Spark's own generated code inside the pipeline rather than falling back to
row-based execution.</li>
+<li><strong>Spark support broadened.</strong> The 0.1.0
release targeted Spark 3.3, 3.4, and 3.5, with experimental 4.0. The
+ 1.0 line drops 3.3, adds 4.1, and ships an experimental 4.2 profile. ANSI
semantics — on by default in
+ Spark 4 — moved from partial to a supported default across the natively
implemented surface.</li>
+<li><strong>The ecosystem story filled in.</strong> Native
Iceberg support arrived in 0.10.0 and has been extended through
+ 1.11 and format V3, native Parquet writes and CSV reads landed as
experimental features and matured,
+ Azure joined S3 on the native cloud path, and 1.0 adds experimental
accelerated PyArrow UDFs alongside the
+ Java and Scala UDF support that shipped in 0.17.0.</li>
+<li><strong>Correctness rigor grew alongside the surface
area.</strong> Early releases relied on fuzz testing to surface
+ divergences; later releases added a full Spark SQL test-suite run against
every supported Spark version,
+ followed by AI-assisted expression audits comparing Comet's behavior against
every supported Spark version
+ edge case by edge case. 1.0 ships with every known open correctness gap
documented in the compatibility
+ guide.</li>
+<li><strong>Performance moved from "modest speedup" to
workload-level wins.</strong> TPC-DS at 1TB has become the reference
+ workload, with successive releases removing FFI round trips, caching Parquet
metadata, shrinking plan
+ serialization, and tightening the shuffle write path. 0.17.0 alone was ~9%
faster than 0.16.0 on TPC-DS
+ 1TB, and the per-expression optimization work in 1.0 continues that
trajectory.</li>
+</ul>
<h2 id="what-10-means">What 1.0 Means<a class="headerlink"
href="#what-10-means" title="Permanent link">¶</a></h2>
-<p>The 1.0.0 release is the culmination of more than two years of work
since the project was <a
href="https://datafusion.apache.org/blog/2024/03/06/comet-donation/">donated</a>
as an
-Apache DataFusion subproject in March 2024, and is less about any single new
feature than about the
-accumulated maturity of the project:</p>
+<p>Beyond the accumulated features, 1.0 is where the project commits to
being something you can build on:</p>
<ul>
<li><strong>Broad Spark coverage.</strong> Comet supports
Apache Spark 3.4.3, 3.5.9, 4.0.4, and 4.1.3 out of the same
codebase, with dedicated Maven profiles, shim sources, and CI matrices for
each, plus an experimental
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
